Our top 10 moments from Snowflake Summit 2025

We put another Snowflake Summit in the books — and what a week it was!

This year’s event was packed with fresh perspectives: we heard a lot about agents (including a thought-provoking conversation with Sam Altman on throwing compute at our toughest business problems); the case for semantic layers got louder; and we took to the keynote stage to share how our long-standing partnership with Snowflake is helping to make Hex’s Magic AI better governed for data teams.

The future’s looking pretty magical! 🔮

From our four incredible speaker sessions (recapped below) to the buzz at Club Hex, to our State of Data Teams Dinner, and Friends of Data event with Snowflake’s Christian Kleinerman, here are some of the highlights of the week.

1. The NBA: How do you measure the ROI of your data team? 🏀

Erik Nylen, the NBA’s AVP, Data Science Lead, shared how he’s evolved their data org with Hex to become a strategic business partner.

💡 Trying to measure the ROI of your data team is a trap. Don’t have a quantifiable ROI number? Don’t stress. The best ROI is about whether your stakeholders are championing your value and impact. Would they recommend you to other teams? Would they give up some of their headcount to the data team? Here’s why measuring data team ROI is a myth.

💡 Want to build relationships? Make speed a priority. "Let me create a ticket for that request and circle back to you in 2 weeks" essentially means "never" to the business stakeholder. Think about how you can speed up answers.

💡 Not all stakeholders care about how the sausage is made. But they want to be convinced that your models are accurate, so set a high standard for scientific excellence.

💡 Politics is part of the data job. The best data teams know how to win friends and influence people — because solving problems isn’t just technical. Data teams should have knowledge of the business challenge that needs to be solved and empathy for how their cross-functional stakeholders need to solve it.

💡 Focus on the new, the novel, and the gnarly. Find a way out of the "qq ad hoc hell" and focus on the thing your data team can uniquely do: tackle your organization's most important, strategic questions. That's what stakeholders will value most.

2. HubSpot: The invisible work behind transforming the data team 🪄

There's an old saying: “If you want to go fast, go alone. If you want to go far, go together.”

Stephen Moseley, Senior Director of Data & Analytics Strategy & Operations, and Anthony Avino, Head of Product & Partner Analytics Engineering, shared how they transformed the data team at HubSpot by creating a better, more collaborative experience — not a bigger stack.

💡 Instead of adding capabilities, they removed friction. They eliminated temporary public Looker dashboards, Looker Explores, and dbt Models by using Hex to explore data faster and auto-archive notebooks after 90 days of inactivity.

💡 Ask yourself, “What’s really slowing us down?” If you find yourself drowning in dashboards and yet still missing insights, the bottleneck is probably not the analysis — it’s likely rooted in siloed tools that make sharing, collaboration, and discovery difficult.

💡 To scale with minimal friction, embrace a mindset of people over tools. It's not just about the tech, data processes run better when you think about people first and tools second. With Hex, HubSpot grows with analysts over time and helps its 8,000+ employees build a culture of learning, transparency, and collaboration from data.

3. Ramp: How do you elevate the influence of your data team? 🏔️

Ian Macomber, Ramp’s Head of Data, shared what it looks like to move up the data science “hierarchy of needs.”

💡 Get your data team to focus on organizational influence. Elevate your data team's influence by having them ask two key questions: "So what, and why?" and "How can I make sure an action is taken based on the results of my work?"

💡 Distribution matters just as much as analysis. Remember to build with the goal of publishing and sharing your work in mind. Ian’s team uses Hex to turn notebooks into interactive data apps that they can embed in Notion.

💡 You need fewer production-grade dashboards than you think. Ian’s team moves fast and iterates quickly by prototyping with Hex, so being more selective of what deserves to be canonical pays off.

💡 Bringing different data personas together in one tool leads to better decision-making. To make more efficient decisions across a growing organization, it’s best to get everyone operating from the same data source. At Ramp, folks over a wide range of technical skills — from applied scientists to analytics engineers to product managers — all collaborate in Hex to arrive at smarter decisions.

4. Chime: We’re in the golden age of the analyst. ⭐

Jordan Farrer, a Director of Data Analytics at Chime, shared how his team uses tools like Hex to change what they spend their brainpower on.

💡 Data tooling is becoming more advanced — look for areas the tools can streamline processes. With high demands on the data team’s time, think critically about processes you can optimize. Lean on your data tool to save time on tasks like getting set up, onboarding, learning IT and security tools, and maintaining data products and data access.

💡 Focus more on how to communicate your work. This includes how to visualize data the right way, share findings, and align on a path forward with stakeholders.

💡 Proactively ask questions instead of reactively answering them. Flip the typical dynamic on its head by prioritizing data products over reactive analysis. That way, your data team can be a true strategic partner — not a dashboard factory.

5. Friends of Data fireside chat with Snowflake’s EVP of Product, Christian Kleinerman ❄️

We were fortunate enough to spend some time with Snowflake’s EVP of Product and ask him some of our burning questions in a fireside chat. The recap is in the works, stay tuned!

6. Hex's co-founder and CTO, Caitlin Colgrove, sharing about Hex’s Magic AI, in the Snowflake keynote 🤖

“At Hex, we’ve been thinking about these multi-step agentic workflows for years now, but prior generations of models really struggled with the complexity and nuance of data work. With these latest generations, including Claude 4, we’ve seen these workflows be unlocked in completely new ways.

We’re very excited to partner with Snowflake to bring the latest generation of agentic tools to data work — helping our users spend less time on rote, repetitive tasks and more time answering the deepest and most impactful questions for their businesses.”

– Caitlin Colgrove, Hex’s Co-founder and CTO

7. The State of Data Teams Dinner 📊

We had a blast launching our State of Data Teams Report at Summit! The results paint an interesting picture of what data teams are thinking about in 2025:

• 77% of data leaders are excited about AI's potential, but only 3% are actively implementing it right now.

• 70% of data leaders describe self-serve as a “worthy goal,” but many have hit roadblocks in making it a reality.

• 84% of data leaders say they spend most of their time thinking about data quality and reliability.

8. Announcing our new integrations with Snowflake Semantic Views and AISQL! 🧩

Semantic models are more relevant than ever in the age of AI. That's why we partnered with Snowflake to introduce direct support for Snowflake’s new Semantic Views inside of Hex.

Now, you can define your semantic models in Snowflake and ingest them seamlessly into Hex — making it faster and easier than ever to deliver consistent, governed insights.

• ✍️ Create Snowflake Semantic Views right from the Hex notebook

• 📊 Jump into Explore for quick and intuitive visual analysis

• 🔮 Use Magic to ask questions on your modeled data

Plus, use Snowflake’s AISQL functions for powerful multimodal LLM capabilities in the Hex notebook.

9. Hex’s co-founder & CEO, Barry McCardel getting interviewed on Data Cloud Now. 📽️

From our State of Data Team report to AI’s impact on data science, Barry covers it all in an interview on Data Cloud Now.

Ryan Green, Snowflake News Anchor: As AI tools become more integrated in everyday workflows, how do you see the roles of data scientists and business users evolving?

Barry McCardel, CEO at Hex: "At a macro-level, there’s just going to be a lot more data use and data consumption. Being able to start from natural language and get an answer is just going to lower that friction enormously. No matter what persona you are, we’re all going to be consuming a lot more data.”

10. Blood Sugar Hex Magic bash ❤️‍🔥

Our community came together for a night of pure magic and mischief! 🪄 The evening was full of mesmerizing magic tricks, clouds of cotton candy, infectious beats from our DJs, and yes — some brave souls even got real tattoos. (Regrets? Never. 😤)

Celebrating our community

Nearly 1,000 of you wonderful data folks dropped by Club Hex, our chill coworking “third space,” during Summit. Whether you geeked out over video games in the Datadog den, shared breakfast with us, grabbed a pick-me-up from our coffee bar, or joined us for some mid-day boba at Jessie Square — we loved meeting each and every one of you. Your passion for data and creative ways of using Hex inspire us every day. Can't wait to see what magic you create next year! ✨